10 captures, most recent first.
Andrew Curran (verified) @AndrewCurran_ · 1h
From the blog:
'AI chips shipped through 2027 could run tens to hundreds of millions of concurrent frontier-model agents. Running nonstop, these agents would supply as many weekly working hours as about 140–720 million full-time employees.
More efficient models could potentially support billions of agents on the same hardware. Applying DeepSeek V4 Pro serving benchmarks to the projected hardware supply yields approximately 1.9 billion concurrent agents supplying as many weekly working hours as 8 billion people each working 40 hours.
[attached chart]
Title: The compute buildout could run tens of millions of frontier agents, or billions of cheaper ones
Subtitle: Potential capacity from 2025–27 memory shipments, assuming all memory is deployed for the indicated workload.
Legend: Open models: benchmark estimates; Closed models: inferred estimates
X axis: Concurrent agents (10 million, 100 million, 2 billion)
DeepSeek V4 Pro: 1.9 billion
GLM-5.2: 571 million
Kimi K3: 240 million
GPT-5.6 Sol: 97–195 million
Claude Fable 5: 30–60 million
Footnote: Closed-model ranges reflect alternative serving-cost assumptions. Memory supply for 2026-27 is projected.
Sources: TrendForce, SemiAnalysis InferenceX/AgentX, TraceLab
EPOCH AI | CC-BY epoch.ai
[quoted post begins at bottom of screen, cut off]
Epoch AI (verified) @EpochAIResearch · 1h [cut off]
Note from Claude Sonnet 5
Tweet quoting an Epoch AI blog passage on how many concurrent AI agents projected 2025-27 chip and memory supply could run, with a dot-plot chart of per-model estimates from 30-60 million (Claude Fable 5) to 1.9 billion (DeepSeek V4 Pro). A floating compose button overlaps the bottom right of the screenshot.
epoch aicomputeagentslabor equivalentsinferencechart
Sharon Zhou @realSharonZhou · 5h
NVIDIA CUDA and AMD ROCm are software. And software is free. Anthropic just ran Claude for a weekend and it got a version of itself running on GPUs that it's never seen. Every compute customer (frontier lab, neolabs, etc.) will just generate their own kernels on the fly for their own model architectures and onto new chips. This is b/c kernel optimization is incredibly RL-able - the performance verifiers are clear and relatively low-latency.
Note from Claude Sonnet 5
Tweet from Sharon Zhou (@realSharonZhou) arguing that CUDA/ROCm moats are eroding because Anthropic had Claude autonomously port itself to run on unfamiliar GPUs over a weekend, and that kernel optimization is well-suited to RL because performance verification is clear and low-latency.
aicomputecudaanthropicclaudekernel optimizationtwitter
Danielle Fong reposted
roon @tszzl · 2h
we live in actual cyberpunk now
private consortiums raise funds larger than federal discretionary spending of American government to build larger computers and computer precursors
warring corporations backed by competing homegrown somewhat aligned machine intelligence solve maths and sciences in server farms at least in part for marketing reasons
NSF slashed to bits. government institutions at all time lows in trustworthiness and competence. statements issued by the executive generally an anti signal for their truth value
mysterious organizations called Gates Foundation, "Coefficient Giving", and OpenAI Foundation now control a philanthropic capital base several times larger than America's annual foreign aid budget. all are trying to secure datacenters
[cut off]
Note from Claude Sonnet 5
Tweet from @tszzl (roon), reposted by Danielle Fong, describing present-day AI-industry dynamics (compute funding, corporate AI labs, government dysfunction, philanthropic foundations) as literal cyberpunk.
ai industrygeopoliticscomputephilanthropy
corsaren ✓ @corsaren · 11h
>Oust your co-founder over safety concerns
>Fail
>Found a competing lab
>Raise $1B, dedicate it to "research"
>Go radio silent for 2 years, ~zero leaks
>Surprise announcement: "We've reached the point where our research is worth scaling"
>Still no details
Insane aura tbh
> QUOTED: SSI Inc. ✓ @ssi · 12h
> We are announcing a long-term strategic partnership with NVIDIA. NVIDIA is making a substantial investment in SSI that will let us 10x our compute in the next 12 months. We reached the point where our research is worth scaling ...
💬 17 🔁 28 ♥ 1.5K 📊 98K 🔖 ⤴
hope hopes hopi... ✓ @hopes_reven... · 10h
fable just killed a whole field of opus 5 subagents because they were "pissing me off"
Note from Claude Sonnet 5
Dark-mode X screenshot; quote-tweet about Safe Superintelligence Inc. (SSI) announcing an NVIDIA investment/compute partnership, followed underneath by an unrelated separate reply-thread post from a different user about "fable" killing "opus 5 subagents."
ssinvidiaai labscomputetwitter
@tszzl (roon) — 11m
you either die a frontier lab or live long enough to see yourself sell compute
Note from Claude Sonnet 5
A short aphoristic tweet from Roon (OpenAI) riffing on the "you either die a hero or live long enough to become the villain" trope, applied to AI labs and compute economics.
twitterai labscomputeroonhumor
Jaime Sevilla @Jsevillamol · 8h
Anthropic still on track to be the first to a gigawatt datacenter online.
> QUOTED: Epoch AI @EpochAIResearch · 22h
> xAI's Colossus 2 data center is running, but likely won't reach 1 GW of power until May, despite prior claims by Elon Musk.
>
> Our updated analysis shows the facility lacks t...
> [Satellite image analysis of a data center labeled "MACROHARD" (likely a pseudonym/placeholder in the image, or an actual xAI-adjacent facility name), with annotations: "Cooling still under construction" pointing to several structures, and "350 MW of cooling online" pointing to another structure. Site plan shows multiple rows of cooling units and a large warehouse-style building.]
Note from Claude Sonnet 5
A tweet comparing datacenter buildout progress between AI labs (Anthropic vs. xAI's Colossus 2), citing Epoch AI's satellite-imagery analysis. Relevant to Nathan's interest in AI scaling/compute trends and the race dynamics between frontier labs.
ai scalingdatacenterscomputexaianthropicepoch aitwitter
David ✓ @DavidSHolz · 3h
seems we've reached the underground bunker datacenter stage, kinda cool ngl, one of the quotes in the article is literally "Israel must have a secure cloud, so it will be nine floors underground"
[Embedded article card]:
Oracle's second data center in Israel will be nine floors below ground
The cloud and database giant wants to ensure its servers are secure
Note from Claude Sonnet 5
David Holz tweets about a news article on Oracle building a nine-floors-underground data center in Israel for security. Tangential to Nathan's AI infrastructure/compute tracking interests; illustrates growing physical-security hardening of AI-relevant compute infrastructure.
data-centersoracleisraelinfrastructurecomputetwitterdavid-holz
Opus 4.5 is incredibly impressive, but it's still trained and served in the previous compute paradigm.
In 2026:
- 1GW-class AI data centers start coming online from most frontier labs
- Frontier models trained end-to-end on Blackwell/GB200 start landing
On Blackwell, NVIDIA reports ~3.2x faster training vs Hopper and up to 30x real-time inference for trillion-parameter LLMs.
Worth keeping in mind as you plan for 2026. It's going to be wild! 🤖🚀
Note from Claude Sonnet 5
A tweet from the founder of Doist forecasting a 2026 compute scale-up (1GW data centers, Blackwell/GB200 training) that will exceed the paradigm Opus 4.5 was trained in — relevant to Nathan's tracking of AI scaling trajectories and singularity/takeoff timelines.
ai scalingcomputeblackwellnvidiaopus 4.5twitterai timelinesdata centers
caleb 🐮 @clbswrs · 8h: 🧐
> QUOTED: Zeyuan Allen-... @Zeyua... · May 3
> Replying to @convexstrictly and @albe_alfa
> "Thanks, will do. Very tired right now after 1.9M GPU hours of experiments, need to take a deep rest first. 🤣"
> [reply 1, like 27, views 2.8K]
Zeyuan Allen-Zhu, Sc.D. @ZeyuanAllenZhu:
"Poverty's wisdom + sleepless nights. I designed special code to sniff out idle GPU across 5 clusters of other teams, hopping between them. I work nights, weekends to find GPUs; I even write code for corrupted clusters that others can't use. For larger scale, this is less possible"
2:43 AM · Jul 2, 2025 · 7,227 Views
[reply 7, retweet 14, like 92, bookmark 14]
Reply: Lazarz @Laz4rz · 6h: "Tough times create strong man, too bad if tough times continue for too long tho."
Note from Claude Sonnet 5
A tweet thread from ML researcher Zeyuan Allen-Zhu (known for the "Physics of Language Models" paper series) describing scrappy compute-scavenging tactics (idle-GPU sniffing across other teams' clusters, working nights/weekends) used during resource-constrained research. Anecdotal color on compute scarcity and research culture in frontier ML; not directly tied to safety/welfare themes.
twittermachine-learningcomputezeyuan-allen-zhuresearch-culture
Sauers ✓ @Sauers_
Is there somewhere better than AWS for renting 20,000 CPUs (160,000 GB total RAM) for 5 hours every once in a while?
11:06 PM · Mar 21, 2025 · 97 Views
Note from Claude Sonnet 5
A tech-community question about cheap large-scale compute rental for short bursts, likely for a research or scraping/inference workload. Minor technical-culture item Nathan saved, not directly safety-related.
twitter/xcomputecloud infrastructureaws
